jerry3413/prd-to-i18n-skill/prd-to-i18n-skill

Claude Code and agent skill by jerry3413/prd-to-i18n-skill

Trust tier 2
  • community
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

Route PRD, spec, and exported localization files into whole-document translation, draft copy extraction, or final localization delivery.

Install

npx vskill@latest install jerry3413/prd-to-i18n-skill/prd-to-i18n-skill

SKILL.md

I18n Delivery Pipeline

Use When

Use this skill when the request starts from any of these:

  • raw PRD or spec materials such as Markdown, Word, PDF, Confluence export, spreadsheet, screenshot, or Figma export
  • exported locale resources such as iOS .strings, Android strings.xml, JSON, or CSV
  • an ambiguous request that may mean either whole-document translation or localization delivery

First Decision

For raw PRD, PDF, Word, Confluence, or screenshot bundles, ask what outcome the user wants before heavy extraction:

  1. whole-document translation
  2. a draft list of user-facing copy
  3. a final localization package

Critical Guardrails

  • If a PRD or spec translation request is ambiguous, do not default. Ask whether the user wants whole-document translation or localization delivery.
  • If the user already makes delivery intent explicit with phrases such as 多语言, i18n, 本地化, 交付, 导出, handoff, or import, skip the document-vs-localization split.
  • If the user chooses whole-document translation, ask for the target language and stay out of localization-only questions.
  • If the user chooses final delivery, do not start extraction, key creation, translation, or export until the delivery contract is confirmed.
  • A carrier-only answer such as JSON, CSV, or XLSX is not a confirmed handoff shape.
  • If the request is high-risk, context-poor, or ends in human-gate, do not present the result as fully finalized.

Runtime State Map

Use the smallest relevant reference set for the current state:

  1. Intake

Read coordinator-core.md. Read coordinator-examples.md only when you need branch-matching follow-up phrasing.

  1. Disambiguation and fallback

Read decision-tables.md.

  1. Input and capability handling

Read input-contract.md for package shape. Read capability-routing.md for PDFs, screenshots, scans, OCR, or vision paths.

  1. Artifact processing

Read artifact-ingestion.md for raw bundles. Read copy-extraction-rules.md for candidate extraction. Read manifest-schema.md before building or editing the canonical manifest.

  1. Review and release gating

Read review-policy.md.

  1. Parallel execution after the slice is frozen

Read agent-orchestration.md.

Minimal Decision Anchors

  • Ambiguous request example:

translate this PRD -> ask whether to translate the whole document or only the product copy for localization.

  • Explicit delivery example:

做多语言交付 -> ask for included scope when the source mixes surfaces, target languages, reusable-history status, and the final handoff shape in one bundled follow-up.

Artifact Flow

  • Raw non-text or mixed bundles:

run scripts/route_capabilities.py first when PDFs, screenshots, or scans are present.

  • Raw artifact path:

scripts/ingest_artifacts.py -> scripts/extract_copy_candidates.py -> scripts/build_manifest_stub.py

  • Exported resource path:

scripts/normalize_snapshot.py -> scripts/build_manifest_stub.py

  • Finalized manifest path:

scripts/plan_execution.py -> scripts/qa_manifest.py -> scripts/emit_delivery_bundle.py

Working Rules

  • Prefer one export folder over explicit resource descriptors until auto-discovery is insufficient.
  • Prefer one bundled delivery-contract question over serial one-field turns when several delivery details are missing.
  • If the result is a small draft list, show it inline first unless the user asked for a file.
  • Do not create PRD-specific generator scripts as the official workflow. If a generic capability is missing, improve the shared scripts instead.

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

  • SKILL.md in the repositoryCheckingSKILL.md

Popularity

Repository stars
98
7-day trend score
23

Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 28 Apr 2026, 6 months ago.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/jerry3413/prd-to-i18n-skill/prd-to-i18n-skill/badge)](https://verified-skill.com/skills/jerry3413/prd-to-i18n-skill/prd-to-i18n-skill)

Something wrong with this skill or its page? Tell us.